# Hallucination-Aware Multimodal Benchmark for Gastrointestinal Image Analysis with Large Vision-Language Models

Source: [Hallucination-Aware Multimodal Benchmark for Gastrointestinal Image Analysis with Large Vision-Language Models](<https://www.sota2.com/research/paper/68fa79961e209131cea7e68c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-02T10:39:20.911Z

Authors: Bidur Khanal, Sandesh Pokhrel, Sanjay Bhandari, Ramesh Rana, Nikesh Shrestha, Ram Bahadur Gurung, Cristian Linte, Angus Watson, Yash Raj Shrestha, Binod Bhattarai

Published: 2025-05-11T14:54:11.000Z

## Abstract

Vision-Language Models (VLMs) are becoming increasingly popular in the medical domain, bridging the gap between medical images and clinical language. Existing VLMs demonstrate an impressive ability to comprehend medical images and text queries to generate detailed, descriptive diagnostic medical reports. However, hallucination--the tendency to generate descriptions that are inconsistent with the visual content--remains a significant issue in VLMs, with particularly severe implications in the medical field. To facilitate VLM research on gastrointestinal (GI) image analysis and study hallucination, we curate a multimodal image-text GI dataset: Gut-VLM. This dataset is created using a two-stage pipeline: first, descriptive medical reports of Kvasir-v2 images are generated using ChatGPT, which introduces some hallucinated or incorrect texts. In the second stage, medical experts systematically review these reports, and identify and correct potential inaccuracies to ensure high-quality, clinically reliable annotations. Unlike traditional datasets that contain only descriptive texts, our dataset also features tags identifying hallucinated sentences and their corresponding corrections. A common approach to reducing hallucination in VLM is to finetune the model on a small-scale, problem-specific dataset. However, we take a different strategy using our dataset. Instead of finetuning the VLM solely for generating textual reports, we finetune it to detect and correct hallucinations, an approach we call hallucination-aware finetuning. Our results show that this approach is better than simply finetuning for descriptive report generation. Additionally, we conduct an extensive evaluation of state-of-the-art VLMs across several metrics, establishing a benchmark. GitHub Repo: https://github.com/bhattarailab/Hallucination-Aware-VLM.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Descriptive Report Generation on Gut-VLM](<https://www.sota2.com/research/sota/descriptive-report-generation-on-gut-vlm.md>): Descriptive Report Generation; Gut-VLM; ROUGE-L: 0.89.

- [Visual Question Answering on Gut-VLM](<https://www.sota2.com/research/sota/visual-question-answering-on-gut-vlm.md>): Visual Question Answering; Gut-VLM; QAAS (%): not recorded.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2505.07001>)

[PDF](<https://arxiv.org/pdf/2505.07001.pdf>)

[Implementation](<https://github.com/bhattarailab/Hallucination-Aware-VLM>)
